el cid again
My question is about how the code SHOULD operate, not a dispute over how it
DOES operate. I once worked on professional simulations for USAF, NASA, etc
at Boeing (at, not for - I was a "resident computer engineer" from a different
defense contractor with my desk at the Boeing Spaceflight Center). I am
confident the description is sincere and correct within your understanding.
[I have correspondence with a chief AE programmer saying that he is never sure
of anything. "After two years working with LCU, I thought I had a firm grasp of
how code works. Today, after finding a branch, I am sure of NOTHING." Emphasis
in the original.]
My point is that this is NOT the best way to model this problem.
That said, AE is almost elegant in the simplicity of its code concepts. And some of
them work astonishingly well. IF one feeds in good aircraft data (which I admit
is hard to do) air combat works better than almost any other model - even those
dedicated to it and far more elaborate). It is impressive. The trick is to figure
out how to define maneuverability and durability in a way that is objectively "fair"
- which is a whole lot easier to say than to do. I am not trying to be overly harsh
about the code. Just saying - a correct approach must define penetration on a
standard basis for ALL weapons. Remember, a heavy gun usually has more range, so it
can shoot at all ranges, while a light gun cannot. As well, the chance of a hit at
long range is very low - it won't happen much. It is much more important to get
penetration right at SHORT range - where it applies to ALL guns - than it is at long
range - where in the nature of things, only long range guns can reach anyway. If one
wanted to be extremely simple, one might define ALL long range hits as "deck armor"
hits - and short and medium range hits as "side armor" hits. But penetration needs to
reflect what happens at short range - where most guns will score most hits. Unless one
writes very sophisticated code - which AE does not go for. Instead, the philosophy was
to use random numbers to reflect other possibilities. Which is NOT a bad philosophy
and certainly is simple. The original programmer loved "die rolls" - and it is often
a very fine mechanism.